Anthropic has released Claude Fable 5, a powerful AI model with built-in safeguards that automatically downgrade sensitive requests in cybersecurity and biology to the less capable Claude Opus 4.8 to prevent misuse. Industry experts are raising concerns about the tiered access model, noting that while a restricted Mythos 5 version is available exclusively to government agencies and select partners, commercial customers pay premium prices but face limitations on security-related queries. Security professionals warn this creates a "security poverty line" where well-funded organizations and attackers gain access to advanced AI capabilities while smaller teams struggle to defend against increasingly automated threats without the same tools.